Autonomous vehicle self-localisation by scene matching under extreme environmental changes has been among the most challenging problems in robotics and computer vision in the last few years. Large dynamic illumination changes during the day and appearance changes between different seasons are the major difficulties about this problem, especially when the comparison is made between daytime and night-time images for the same scene. This study presents a new extended local difference binary (ELDB) image descriptor that represents a robust appearance invariant extension for the state-of-the-art local difference binary (LDB) image descriptor. This study also introduces a new algorithm for vehicle visual localisation under extreme environmental changes. The new algorithm uses ELDB for image matching, and uses a modified multi-hypothesis version of the Markov localisation (MHML) filter for self-localisation. Experimental results show that the proposed modified MHML has reduced computational cost and has resulted in a faster cycle rate. Furthermore, these results show that ELDB has an improved image matching accuracy and requires less processing time compared to the original LDB. The proposed visionbased vehicle localisation algorithm is shown to be faster and more accurate than other state-of-the-art algorithms.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.